Abstract. The effectiveness of animation in visualization is an interesting research topic that led to contradicting results in the past. On top of that, we are facing three additional challenges when exploring patient cohorts: irregular sampling, data wear, and data sets covering different portions of time. We present TimeRider, an improved animated scatter plot for cohorts of diabetes patients that tackles these challenges along with its evaluation with physicians. Results show that animation does support physicians in their work and provide further domain-specific evidence in the discussion on the effectiveness of animation.